RNA interference refers to a mechanism, which is post-transcriptional gene silencing initiated by a double-stranded RNA (dsRNA) via nucleotide sequence specific manner in a gene expression process, and this mechanism is first found in C. elegans, and commonly found in plant, fruitfly, and vertebrate (Fire et al., Nature, 391:806-811, 1998; Novina & Sharp, Nature, 430:161-164, 2004). It has been known that RNA interference occurs in such a manner that dsRNA of 19-25 bp entering in the cell is bound with an RISC (RNA-induced silencing complex), and only an antisense (guide) strand is bound with mRNA such that it is complementary to the nucelotide sequence of the mRNA, thereby degrading target mRNA by endonuclease domains existing in the RICS (Rana, T. M., Nat. Rev. Mol. Cell Biol., 8:23-36, 2007; Tomari, Y. and Zamore, P. D., Genes Dev., 19: 517-529, 2005).
When the dsRNA is delivered into a cell, it is specifically bound to a target mRNA sequence to degrade the mRNA, and thereby, it is considered as a new tool capable of regulating gene expression. However, in case of human, it was difficult to obtain RNAi effect due to the induction of an antiviral interferon pathway on introduction of dsRNA into human cells. In 2001, Elbashir and Tuschl et al., found that the introduction of small dsRNA of 21nt length (nucleotides length) into human cells did not cause the interferon pathway but specifically degraded the target mRNA (Elbashir, S. M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K., Tuschl, T., Nature, 411, 494-498, 2001; Elbashir, S. M., Lendeckel, W., Tuschl, T., Genes & Dev., 15, 188-200, 2001; Elbashir, S. M., Martinez, J., Patkaniowska, A., Lendeckel, W., Tuschl, T., EMBO J., 20, 6877-6888, 2001). Thereafter, dsRNA of 21nt length has been spotlighted as a tool of new functional genomics and named as small interfering RNA (siRNA).
The siRNA is a substance gaining a lot of interest as a agent for gene therapy ever since it was reported to have an excellent effect in inhibiting expression of a specific gene in animal cells. In effect, because of its high activity and precise gene selectivity, siRNA is expected to be an alternative therapeutic agent to an antisense oligonucleotide (ODN) currently being used as a therapeutic agent, as a result of a 20-year research (Dana J. Gary et al. Journal of Controlled Release 121:64-73, 2007). A siRNA technique aiming to therapy has great advantages in that it is easily designed compared with other medicines and has high target selectivity and a property of inhibiting expression of a specific gene. In addition, it is less toxic because RNA interference suppresses gene expression by using a mechanism naturally existing in a living system. ‘Bevasiranib’, recently developed as a therapeutic agent for wet age-related macular disease by OPKO Inc., is a siRNA which acts selectively on a vascular endotherial growth factor (VEGF) inducing neovascularization to inhibit expression of the VEGF, and passes through three phases of clinical trial (Dejneka N S et al., Mol Vis., 28(14):997-1005, 2008). Besides, therapeutic agents including siRNAs targeting various genes are currently being developed (Ryan P. Million, Nature Reviews Drug Discovery 7: 115-116, 2008).
Despite various results showing that specific expression inhibition is induced in vivo through RNA interference, in vivo siRNA delivery has many problems to be solved, such as degradation by enzymes in the blood, interaction with components in the blood, and non-specific delivery to cells (Shigeru Kawakami and Mitsuru Hashida, Drug Metab. Pharmacokinet. 22(3): 142-151, 2007). Attempts to overcome these problems are in progress by partially using nuclease resistant nucleoside analogues or improving delivery techniques.
Examples of the improved delivery techniques include gene delivery techniques using viruses such as adenoviruses, retroviruses, etc., and gene delivery techniques by non-viral vectors using liposomes, cationic lipid, and cationic polymer compounds. However, viral carriers has a problem in safety since delivered genes are likely to be integrated into a chromosome of a host to induce abnormality in normal functions of genes of the host and activate oncogenes, and in addition, may cause autoimmune diseases due to successive expression of viral genes even in small amounts, or may not lead to efficient protective immunity in a case where modified viral infection is induced from the viral carriers. Meanwhile, non-viral carriers are less efficient than the viral carriers, but have advantages of low side effects and inexpensive production costs, considering in vivo safety and economic feasibility (Lehrman S., Nature. 401(6753): 517-518, 1999). In addition, non-viral delivery methods require to effectively protect enzymatic or non-enzymatic degradation in order to deliver RNA molecules including siRNA, one method of which is to utilize DNA expression plasmids encoding a short hairpin RNA (shRNA). A system through DNA has an advantage in that siRNA is expressed only while an expression vector exists. Moreover, a recent study on chemical modification of siRNA has proposed a method for improving the stability against nucleases and the low intracellular uptake (Shigery Kawakami and Mitsuru Hashida. Drug Metab. Parmacokinet. 22(3): 142-151, 2007).
In one type of chemical modification of siRNA, a phosphorodiester bond, which is a part degraded by the nuclease, was modified with a phosphorothioate linkage or the 2′ portion of a pentose is modified with 2′-O-meRNA, 2′-deoxy-2′-fluouridine, or a locked nucleic acid (LNA) formed by linking the 2′ portion and the 4′ portion, and as a result, the stability in the serum was improved ((Braasch D. A. et al. Bioorg. Med. Chem. Lett. 14:1139-1143, 2003; Chiu Y. L. and Rana T. M., RNA, 9:1034-1048, 2003; Amarzguioui M. et al. Nucleic Acid Res. 31:589-595, 2003). In another type of chemical modification, a functional group is linked to a 3′-end region of a sense (anti-guide) strand, resulting in improvement in pharmacokinetic characteristics compared with a control, and high efficiency is induced at the time of application in vivo through a balance between hydrophilicity and hydrophobilicity of siRNA (Soutschek J. et al. Nature 432:173-178 2004).
However, the above methods still leave much to be desired in order to protect siRNA from nucleases and improve the efficiency of cell-membrane permeability.
For that reason, the inventors have found that a conjugate, in which hydrophilic or hydrophobic polymer compound is conjugated to siRNA by using a degradable or a non-degradable bond, improved in vivo stability of siRNA, and, based on this, has completed the present invention.